COMP0035 2023/24 Coursework 1 specification
Version: Approved. Version 1.0 27/09/23
Contents
Introduction
Getting started
Technologies to use
Coursework content
Submission
Marking
Appendix
Introduction
This document specifies coursework 1 which is worth 50% of the assessment marks available for
the module.
The aim of this coursework is for you to select and apply relevant software development and
data science techniques for aspects of a project lifecycle to define and design two data science
web applications. You are also expected to evidence the use of appropriate software engineering
tools and techniques in your work.
This assessment includes:
Data preparation and understanding
Product and project definition
Tools and techniques
Methodology (groups only)
You will submit a portfolio of work with evidence for each section. The required evidence is
detailed in section 5 Coursework content.
Coursework 2 will continue from coursework 1, covering further aspects of the lifecycle.
For students continuing to COMP0034 Application Programming for Data Science, in
COMP0034 you will develop and test the applications that you design in COMP0035.
Getting started
There are 3 steps in this section:
1. Choose to work as an individual or group
2. Be allocated a data set
3. Accept a GitHub classroom assignment
Step 1. Choose to work as an individual or group
Please carefully consider the choice as you cannot change your decision during the course i.e. if
you work as a group for coursework 1 you must also work with the same group for coursework 2
(likewise for individuals).
Groups should carefully consider the potential for conflicting timetables and competing time
demands in your 3rd year which can make group work more challenging.
If you choose to work as a group, read thegroup work guidance in Moodle.
Use the selection task on Moodle to register your choice.
Step 2. Be allocated a data set
Data sets are allocated as soon as the group selection task is completed, typically at the end of
the first week of the module.
Where multiple groups/individuals are allocated the same dataset, ensure that you do not
collude with other students ; collusion is considered Academic Misconduct. It is expected that
those using the same dataset will have very different coursework as you will each have your
own product ideas, target audience, questions etc.
Step 3. Accept a GitHub classroom assignment
Create a new repository by accepting the GitHub classroom assignment. Instructions are also
given in the first tutorial.
Individuals
1. Login to GitHub.com.
2. Click on the GitHub classroom link for individuals
3. Accept the assignment.
4. If prompted, accept to join the comp0035-ucl organisation.
Groups
1. Login to GitHub.com.
2. Click on the GitHub classroom link for groups
3. One person only (first person from the group)
1. Accept the assignment.
2. If prompted, accept to join the comp0035-ucl organisation.
3. Create the group (team) using the same name as your Moodle group name
(e.g. Group1, Group2 etc.).
4. All other group members
1. Accept the assignment.
2. If prompted, accept to join the comp0035-ucl organisation.
3. Select your group (team) name from the list.
Technologies to use
Technologies that must be used
Technology Notes
GitHub for
source code
control
All elements of the coursework are to be maintained using source code
control.
Source code control should be in GitHub, with repositories created in
GitHub classroom to allow tutors and PGTAs to gain access.
If you can’t use GitHub for some reason please contact the course tutor to
agree alternative source code control.
Python
coding
environment
You need a python coding environment. Guidance can be found on Moodle.
PDF or
Markdown
for text
1
Text created using any software such as Word, Powerpoint, Excel etc must
be saved as .pdf. Markdown is a widely used format in software
development so you may use this. GitHub provides a markdown guide
though there are many freely available guides online.
pandas and
matplotlib
for data
preparation
You must use these libraries for the data preparation and exploration. Do
not use additional libraries unless absolutely necessary (i.e. if pandas or
matplotlib do not have the necessary capability).
1 Students who prefer audio to text may audio record any textual submissions instead. Use
.mp3 format. Do not create video as you can be more easily identified.
Use of AI
You are permitted, but not required, to use AI (e.g. GitHub Copilot, ChatGPT)only for the
data preparation code.
You are not permitted to use AI for any other section (including the explanation of your
data preparation code).
You must state and explain your use of AI in the ‘tools and techniques’ section of the course
content.
Coursework content
Creating the content
The following subsections state what must be produced. You can choose to provide any
additional relevant items that you believe are useful.
Variances for groups and individuals are noted within each section.
Use techniques included in the teaching materials, or other techniques that are typically used
in software engineering and data science. Please provide a relevant reference if you use a
technique not included in the course materials.
Add all the content to your GitHub repository.
Add all the written content is to be added to the same file in the order shown below for
submission. Please use the section numbering shown below:
1. Data preparation and understanding
1.4 Explanation of code preparation & understanding
2. Product and project definition
2.1 Problem statement
2.2 Product overview
2.3 Persona
2.4 Project goal & objectives / Questions
3. Tools & techniques
3.1 Source code control
3.2 Linting
3.3 Project planning and tracking
3.4 Use of AI
4. Methodology
4.1 Methodology selection
5. References
Do not change the names of the files given in the starter code and use the
filenames given in the following sections. Automated scripts are used to separate files for
marking purposes, if you change the filenames your coursework is likely to be missed from
marking.
1. Data preparation and understanding
The purpose of this section is:
to demonstrate that you understand the data. You should have a good understanding of
the data set structure, attributes and their data types, statistics, distribution etc.
to prepare the data for later use in developing the web apps (products). For example,
ensure you have the data that you need, remove any unnecessary data, address any data
quality issues you discover (e.g. missing values, inappropriate data types), etc.
For this section add evidence for the following to the indicated files:
Item Individual Group Filename
1.1 Python code to
prepare/understand the data
✓ ✓ data.py
1.2 Original data set ✓ ✓ dataset.(csv/xlsx)
1.3 Prepared data set ✓ ✓
dataset_prepared.
(csv/xlsx)
1.4 Explanation of code preparation
& understanding
✓ ✓
coursework1.
(pdf/md/mp3)
Code and data set
Write python code that makes use of the pandas and matplotlib libraries to explore and prepare
the data.
Include the original dataset and the dataset after the code has been applied to it.
You may create charts as part of your preparation and understanding e.g. to examine
distributions, outliers etc. Do not spend significant effort formatting any charts you create as
the visual aesthetic is not assessed.
You may have a data set that is already well-prepared with few corrections or changes needed.
You should still demonstrate in the code that you have examined the data set to show that this
is the case, and you should still explain your findings.
Groups are expected to spend more time than individuals on the code. This is because all
group members must understand the code that is created.
Code quality will be considered in the marking. This includes:
code structure (e.g. use of functions and/or classes)
adherence to appropriate python conventions (e.g., PEP8 style guide , PEP275 docstring
conventions)
documentation (comments, docstrings) and more generally whether your code is easy for
others (in this case the marker) to read and understand.
Code quality will not consider file length. Typically, code that exceeds 300-400 lines of code
would be split into modules. For this coursework, please ignore this aspect of code quality as for
marking purposes all the code must be in a single python file.
Explanation
Describe and explain the results of your code to prepare and explore the data. Explain any
decisions that you took.
For example, if you identify code quality issues then explain how you addressed these; or if you
chose not to address them, then explain the reason for not addressing them.
There is no word limit for this as it depends on the format used and the number of decisions
made.
2. Product and project definition
The purpose of this section is to provide those working on, or those interested in, the project
with an understanding of what the project will deliver (i.e. the web apps products); why these
apps are needed; and who they are being created for.
For this section add all the following to a single file named coursework1.pdf/md:
Item Individual Group Filename
2.1 Problem statement ✗ ✓ coursework1.pdf/md
2.2 Product overview ✓ ✓ coursework1.pdf/md
2.3 Persona ✓ ✓ coursework1.pdf/md
2.4 Project goal & objectives
or Questions
✗ ✓ coursework1.pdf/md
2.1 Problem statement
Write an appropriately structured problem statement, or a product vision, that summarises the
problem or opportunity your web app product (App2) will attempt to provide a solution to.
2.2 Product overview
You will be designing two web app products in coursework 2. The product overview will be
very short, likely a few sentences. It should make clear the type of web product (web app) that
you plan to develop and the focus of the data. There is no formal technique or template for this.
It provides a context for the marker.
The two web apps products are:
App1: A REST API that will allow developers to programmatically access and use your
dataset in their apps. The target audience for this is software developers.
App2: A web app product that makes use of your dataset via the REST API (App1).
You can choose from the following types of app:
a. A data visualisation / dashboard app
b. An app that deploys a machine learning model
c. A web app that uses the data in some other way e.g. to generate content that
has interactive features
See Moodle for links to real world apps that are examples of 2a and 2b. It is expected most
students will create an app of type 2a or 2b. 2c is intended for those who can imagine an app
using the data that doesn’t fit either a. or b.
2.3 Persona
Create a persona(s) that describes the target audience(s) for App2 only. A single persona is
typical though there may be instances where you intend your web app to be used by multiple
personas in which case you can create as many as needed. Creating more personas is not an
indicator for higher marks.
For App1 the target audience will be software developers who need to use your REST API in
their own web app. You do not need to create a persona for this target audience.
2.4 EITHER ‘goals and objectives’ OR ‘questions’
Provide either option 1 or option 2. Providing both will not lead to more marks.
Option 1: Write project goal(s) and objectives. Objectives should be SMART. These can focus
on just App2 or both apps if you prefer.
Option 2 (for data visualisation apps only ): Write questions you will attempt to answer by
creating visualisations ( charts) using your dataset. The questions should relate to the target
audience and the problem statement.
3. Tools and techniques
The purpose of this section is to demonstrate appropriate and effective use of relevant software
engineering tools and techniques.
For this section add evidence for all the following to coursework1.pdf/md:
Item Individual Group Filename
3.1 Source code control ✓ ✓ coursework1.pdf/md
3.2 Linting ✗ ✓ coursework1.pdf/md
3.3 Project planning and tracking ✗ ✓ coursework1.pdf/md
3.4 Use of AI ✓ ✓ coursework1.pdf/md
3.1 Source code control
Add the URL for your repository to coursework1.(pdf/md).
Make regular use of source code control throughout the project. Use it for all the coursework
documents, not just code.
For groups all members are expected to show evidence of source code control. This can be seen
in the commit history.
3.2 Linting
Make use of a linter to demonstrate how effectively your code meets Python style standards
such as PEP8, PEP257.
In coursework1.(pdf/md):
provide evidence of the results of running the linter.
explain how you used this to improve your code quality.
state which Python linter you used.
3.3 Project planning and tracking (groups only)
Provide evidence of regular project planning and tracking incoursework1.(pdf/md).
These should indicate what all members of the group did (use initials rather than full names). A
typical format might be similar to the Scrum daily stand up:
What did you do in the last reporting period?
What do you plan to do in the next reporting period?
State any issues that are blocking progress
You may use any planning and tracking technique and tool that you think is relevant. Options
include but not limited to:
Add .txt or .md files to your GitHub with your weekly plan/progress/issues
Submit reports using the optional weekly Moodle reporting task (see Further resources
section for each week)
Use GitHub Projects to add and assign tasks
Past students have also used Kanban-style boards, Sprint plans/reviews, etc. When using
external tools you must save the evidence (e.g. to PDF) and include it in your submission. A
URL hyperlink to an external tool is not acceptable for marking.
The exact timing of the report is not graded; though reports that are submitted just before the
assessment deadline or where there is no evidence of timing are not considered to as ‘regular’
reporting.
3.4 Use of AI
You are permitted to use code assistance AI (e.g. GitHub Copilot, ChatGPT)only for the data
preparation code.
If you did not use AI, then please clearly state “AI not used” in this section.
If you did use AI then explain your use of AI in this section. For example, state which tool(s)
you used then explain how you used them and to what extent the resulting code was used.
Note, if you do not explain your use of AI here, and it is found to have been used in the
coursework, then it may be considered as plagiarism.
4. Methodology
The purpose of this section is to demonstrate that you can select an appropriate methodology
for a given situation. This is one of the module learning outcomes.
For this section add the following to coursework1.(pdf/md):
Item Individual Group Filename
4.1 Methodology selection ✗ ✓ coursework1.pdf/md
Select a methodology, or combination of methodologies, that is suitable for use in the
project.
Justify your choice. For example, you could define selection criteria and review how well
your choice meets these. Use relevant literature to support your views and (if relevant) your
own experience.
There is no ‘correct’ choice, what is important is that you demonstrate why what you have
chosen is relevant to the particular nature and circumstances of your project.
You can consider methodologies from data science and software engineering.
The ‘project’ in this instance refers to all the activities that you will carry out across both
coursework 1 and 2 in both COMP0035 and COMP0034. The activities will include project and
product definition, data preparation, requirements analysis, application design, development,
testing.
There is no word limit for this as it depends on the format used to present your response. For
example, some use tables, others write textual paragraphs that tend to be longer. There are
past examples on Moodle.
5. References
The methodology section is expected to have some references.
Other sections are a report of your work, rather than a dissertation. As a result you are
unlikely to have many references.
If you do have any then please state them here.
5. Submission
Please refer to Moodle for the deadline date and time.
Submit your work on Moodle as a single .zip in the assignment submission (see Checklist in the
next section). Moodle is used as the submission date/time and to authenticate students. GitHub
is NOT an acceptable alternative for submission.
Check carefully that you have the correct files with the correct file names. Automated scripts
will be used to separate elements for marking; if you change the file name your work is likely to
be missed.
Files that are only provided as URLs that link to external locations will be excluded from
marking consideration since they may be modified after the coursework is submitted.
GitHub provides an option to download repository contents as a zip.
You can include other files in the zip; though marking will only consider the files stated in the
coursework content sections above.
DO NOT add your .venv folder to the zip file, this creates unnecessarily large zip files.
Submission checklist
Check you have the following content in each file:
Section Filename Item Individual Group
1. Data
preparation &
understanding
data.py
1.1 Python code to
prepare/understand the
data
✓ ✓
dataset.(csv/xlsx) 1.2 Original dataset ✓ ✓
dataset_prepared.
(csv/xlsx)
1.3 Prepared dataset ✓ ✓
coursework1.
(pdf/md/mp3)
1.4 Explanation of code
preparation &
understanding
✓ ✓
2.
Product/project
definition
coursework1.
(pdf/md/mp3)
2.1 Problem statement ✗ ✓
2.2 Product overview ✓ ✓
2.3 Persona ✓ ✓
2.4 Project goal &
objectives
or Questions
✗ ✓
3. Tools &
techniques
coursework1.
(pdf/md/mp3)
3.1 Source code control ✓ ✓
3.2 Linting ✗ ✓
3.3 Project planning and
tracking
✗ ✓
3.4 Use of AI ✓ ✓
4. Methodology
coursework1.
(pdf/md/mp3)
4.1 Methodology selection ✗ ✓
6. Marking
Mark allocation
Marks will be allocated for the coursework as follows.
It is expected each person will spend 25 hours on the coursework during the first 5 weeks of the
module and part of reading week.
Groups will have 100 hours in total, 76 hours to complete the deliverables and 24 hours coordination overhead. An overhead for co-ordination is expected for groups. This has been
assumed at 1 hour per person per week for 6 weeks ( total 24 hours). This allows for a 30-
minute weekly meeting as well as review of work items.
While not exact, an indication of the hours of effort given the marking weighting is shown
below.
Section Group Individual
Data preparation and understanding 30% (23 hours) 60% (15 hours)
Product/project definition 20% (15 hours) 20% (5 hours)
Tools and techniques 20% (15 hours) 20% (5 hours)
Methodology 30% (23 hours) -
Mark calculation
Marks for individuals will be given for each section. Each section is then weighted using the %
in the table above to calculate the mark.
A group’s raw mark is calculated using the same method as for individuals. All members of a
group are required to complete an IPAC assessment that includes their own and peers
performance. The calculated IPAC rating will then be applied to the group raw mark to
calculate the specific mark for each person within the group, i.e.
individual group member mark = group raw mark * individual's IPAC rating
IPAC guidance is given in the appendix .
Grading criteria
The coursework will be assessed according to the standards set in the standard UCL Computer
Science grading criteria ( see copy on Moodle in the Assessment section).
The criteria most relevant to this assessment are:
1 Quality of the response to the task set: answer, structure and conclusions
2 Understanding of relevant issues
3 Engagement with related work, literature and earlier solutions
4 Analysis: reflection, discussion, limitations
5 Algorithms and/or technical solution (note: this applies to the code aspects only)
8 Writing, communication and documentation
9 Formatting aspects, visuals, clarity, references
The following tables gives indicators that are more specific to this coursework. Given the openended nature of the coursework it is not possible to describe every possible aspect that could be
considered for each level. Please use the tables below as a guide and not an absolute.
Some aspects relate only to groups, individuals, or both. Check ‘section 5 coursework content’
for what each needs to include.
NOTE: Marks are not awarded for doing more. For example, individuals choosing to do
items identified for groups will not achieve higher marks as this would be unfair since groups do
not have the same opportunity.
Data preparation and understanding
Grade
band
Generic CS Descriptor Coursework-specific indicators
Distinction
90+
Exceptional response with a
convincing, sophisticated argument
with precise conclusions.
Exceptional grasp of complexities and
significance of issues.
Exceptional thought and awareness of
relevant issues. Sophisticated sense of
conceptual framework in context.
Distinction
70-89
A distinctive response that develops a
clear argument and sensible
conclusions, with evidence of nuance.
Thorough grasp of issues; some
sophisticated insights.
Evidence of innovative analysis.
Concepts deftly defined and used with
some sense of theoretical context.
There is evidence of a thorough
understanding of the data. Analysis
and preparation is clearly explained
and decisions justified.
Code quality is high with effective
use of functions or classes, and
adheres to relevant Python
standards (e.g. PEP8, PEP257).
Merit 60-
69
A sound response with a reasonable
argument and straightforward
conclusions, logical conclusions.
Sound understanding of issues, with
insights into broader implications.
There is evidence of a good
understanding of the data. Steps
have been identified that allow the
data set to be immediately useful in
the development phase of the
project.
Code quality mostly adheres to
Python standards.
High pass
50-59
A reasonable response with a limited
sense of argument and partial
conclusions.
Reasonable grasp of the issues and
their broader implications.
Reasonable reproduction of ideas from
taught materials. Rudimentary
definition and use of concepts.
There may be limited evidence that
the student has understood and/or
prepared the data. Any decisions
taken are not clearly explained.
Reasonable code, using basic
required concepts, several flaws in
implementation.
Low pass
40-49
An indirect response to the task set,
towards a relevant argument and
conclusions.
Rudimentary, intermittent grasp of
issues with confusions.
Analysis relying on the partial
reproduction of ideas from taught
materials. Some concepts absent or
wrongly used.
Insufficient evidence that the data
has been understood.
Rudimentary preparation code, but
mostly incomplete.
Grade
band
Generic CS Descriptor Coursework-specific indicators
Product and project overview
Grade
band
Generic CS Descriptor Coursework-specific indicators
Distinction
90+
Exceptional response with a
convincing, sophisticated argument
with precise conclusions.
Exceptional grasp of complexities and
significance of issues.
Expert-level review and innovative
synthesis (to a standard of academic
publications).
Exceptional thought and awareness of
relevant issues. Sophisticated sense of
conceptual framework in context.
Reads as if professionally copy edited.
Distinction
70-89
A distinctive response that develops a
clear argument and sensible
conclusions, with evidence of nuance.
Thorough grasp of issues; some
sophisticated insights.
Extensive reading and thorough
understanding of literature consulted.
Excellent critical analysis of literature.
Evidence of innovative analysis.
Concepts deftly defined and used with
some sense of theoretical context.
Style and word choice show fluency
with ideas and excellent
communication skills.
Referencing is impeccable.
A range of relevant artefacts are
presented.
The problem is clearly defined with
relevant information.
The target audience is described
using an appropriate format and
with a level of detail that will
support later activities concerned
with the dashboard and app design.
Questions to be addressed by
analysis of the data set have been
clearly formulated and articulated
and demonstrate insight. They are
clearly linked to both the problem
and audience.
Merit 60-
69
A sound response with a reasonable
argument and straightforward
conclusions, logical conclusions.
Sound understanding of issues, with
insights into broader implications.
Evidence of plentiful relevant reading
and sound understanding of literature
consulted.
Evidence of student’s own analysis.
Concepts defined and used
systematically/ effectively.
Style and word choice work well to
convey most important ideas. Well
documented.
Consistent referencing.
The problem and target audience
are clearly defined and are provided
in a format that is relevant for the
nature of the project.
Questions to be answered by the
analysis of the dataset are clearly
given but may show minor
inconsistency with the data set
provided.
High pass
50-59
A reasonable response with a limited
sense of argument and partial
conclusions.
Reasonable grasp of the issues and
their broader implications.
Evidence of relevant reading and some
understanding of literature consulted.
Reasonable reproduction of ideas from
taught materials. Rudimentary
definition and use of concepts.
The problem statement is stated
but may lack some clarity.
The description of the target
audience may be superficial and/or
of a format that is not specific to
the nature of the project. The
questions to be answered by the
data analysis may be unclear or
inconsistent with the data set.
Grade
band
Generic CS Descriptor Coursework-specific indicators
Low pass
40-49
An indirect response to the task set,
towards a relevant argument and
conclusions.
Rudimentary, intermittent grasp of
issues with confusions.
Significant omissions in reading with
weak understanding of literature
consulted.
Analysis relying on the partial
reproduction of ideas from taught
materials. Some concepts absent or
wrongly used.
Style and word choice seriously detract
from conveying of ideas.
Grade
band
Generic CS Descriptor Coursework-specific indicators
Methodology
Grade
band
Generic CS Descriptor
Coursework-specific
indicators
Distinction
90+
Exceptional response with a convincing,
sophisticated argument with precise conclusions.
Exceptional grasp of complexities and significance
of issues.
Expert-level review and innovative synthesis (to a
standard of academic publications).
Exceptional thought and awareness of relevant
issues. Sophisticated sense of conceptual
framework in context.
Reads as if professionally copy edited.
Distinction
70-89
A distinctive response that develops a clear
argument and sensible conclusions, with evidence
of nuance.
Thorough grasp of issues; some sophisticated
insights.
Extensive reading and thorough understanding of
literature consulted. Excellent critical analysis of
literature.
Evidence of innovative analysis. Concepts deftly
defined and used with some sense of theoretical
context.
Style and word choice show fluency with ideas
and excellent communication skills.
Referencing is impeccable.
Choice of methodology is
clearly explained and
justified within the
context of the project and
shows evidence of nuance
in understanding the
issues involved in
selecting and using an
appropriate method.
The discussion is
supported by relevant
references.
Merit 60-
69
A sound response with a reasonable argument
and straightforward conclusions, logical
conclusions.
Sound understanding of issues, with insights into
broader implications.
Evidence of plentiful relevant reading and sound
understanding of literature consulted.
Evidence of student’s own analysis. Concepts
defined and used systematically/ effectively.
Style and word choice work well to convey most
important ideas. Well documented.
Consistent referencing.
A methodology has been
selected that is
appropriate for the
project proposed.
High pass
50-59
A reasonable response with a limited sense of
argument and partial conclusions.
Reasonable grasp of the issues and their broader
implications.
Evidence of relevant reading and some
understanding of literature consulted.
Reasonable reproduction of ideas from taught
materials. Rudimentary definition and use of
concepts.
A methodology has been
selected though it may
not be clear why this is
appropriate for the given
project.
Grade
band
Generic CS Descriptor
Coursework-specific
indicators
Low pass
40-49
An indirect response to the task set, towards a
relevant argument and conclusions.
Rudimentary, intermittent grasp of issues with
confusions.
Significant omissions in reading with weak
understanding of literature consulted.
Analysis relying on the partial reproduction of
ideas from taught materials. Some concepts
absent or wrongly used.
Style and word choice seriously detract from
conveying of ideas.
Methodology explanation
is limited and/or the
selected methodology is
inappropriate for the
nature of the project.
Grade
band
Generic CS Descriptor
Coursework-specific
indicators
Software engineering tools and techniques
Source code control will be assessed using the commit history in the repository. You must
include the URL in your submitted work.
There are no generic CS criteria relating to this aspect.
Grade
band
Coursework-specific indicators
Distinction
90+
Near flawless use of a range of appropriate tools and techniques. Likely to be
evidence that tools not taught in the course have been carefully selected and
used.
Distinction
70-89
Exceptional use of the expected tools and techniques.
Merit 60-
69
Effective use of the expected tools and techniques.
High pass
50-59
Appropriate use of the expected tools and techniques, those the use of these
may be less effective.
Low pass
40-49
Limited, or inappropriate, use of the expected tools and techniques.
The following questions are not exhaustive and are indicators only:
Source code control
Appropriate:
has it been used regularly? i.e. not just for a day or two before the coursework
deadline
has it been used by all team members? Everyone needs to demonstrate competence
Effective:
has it been used effectively? e.g. are the commit messages clear and specific?
are there typical files used appropriately e.g. README.md, .gitignore? Are
appropriate files excluded e.g. virtual environments?
Exceptional: is there any evidence of sophistication? e.g. branches, pull requests, other
features of GitHub
Linting
Limited: has a linter been used? e.g. is there eviden